Research article

Biogeosciences, 19, 2741–2757, 2022

https://doi.org/10.5194/bg-19-2741-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Modeling interactions between tides, storm surges, and river

discharges in the Kapuas River delta
Joko Sampurno1,2 , Valentin Vallaeys1 , Randy Ardianto3 , and Emmanuel Hanert1,4
1 Earth and Life Institute (ELI), Université catholique de Louvain (UCLouvain), Louvain-la-Neuve, 1348, Belgium
2 Department of Physics, Fakultas MIPA, Universitas Tanjungpura, Pontianak, 78124, Indonesia
3 Pontianak Maritime Meteorological Station, Pontianak, 78111, Indonesia
4 Institute of Mechanics, Materials and Civil Engineering (iMMC), Université catholique de Louvain (UCLouvain),

Louvain-la-Neuve, 1348, Belgium

Correspondence: Joko Sampurno (joko.sampurno@uclouvain.be, jokosampurno@physics.untan.ac.id)

Received: 19 October 2021 – Discussion started: 4 November 2021

Revised: 29 April 2022 – Accepted: 9 May 2022 – Published: 1 June 2022

Abstract. The Kapuas River delta is a unique estuary sys- flooding event in Pontianak, which occurred on 29 Decem-
tem on the western coast of the island of Borneo, Indone- ber 2018. For this event, the wind-generated surge appeared
sia. Its hydrodynamics are driven by an interplay between to be the dominant trigger.
storm surges, tides, and river discharges. These interactions
are likely to be exacerbated by global warming, leading to
more frequent compound flooding in the area. The mech-
anisms driving compound flooding events in the Kapuas 1 Introduction
River delta remain, however, poorly known. Here we at-
tempt to fill this gap by assessing the interactions between Global warming leads to more frequent tropical storms, ris-
river discharges, tides, and storm surges and how they can ing sea levels, and more intense rainfalls, which all concur
drive a compound inundation over the riverbanks, partic- in increasing the occurrence of compound flooding events
ularly within Pontianak, the main city along the Kapuas in a coastal area and its surrounding environment (Bevac-
River. We simulated these interactions using the multi-scale qua et al., 2020). Such disasters impact the lives of more
hydrodynamic model SLIM (Second-generation Louvain-la- and more people worldwide (Moftakhari et al., 2015). Com-
Neuve Ice-ocean Model). Our model correctly reproduces pound flooding occurs when dry low-lying land (over a
the Kapuas River’s hydrodynamics and its interactions with coastal area) is flooded by water from the ocean and the river.
tides and storm surge from the Karimata Strait. We consid- Compound flooding is driven by the interaction between a
ered several extreme-scenario test cases to evaluate the im- coastal inundation and a riverine inundation. The former hap-
pact of tide–storm–discharge interactions on the maximum pens from the seaward direction due to high (spring) tides
water level profile from the river mouth to the upstream part and storm surges, while the latter occurs from the landward
of the river. Based on the maximum water level profiles, due to high discharges from upstream rivers. Coastal com-
we divide the Kapuas River’s stream into three zones, i.e., munities, which have been growing in population over the
the tidally dominated region (from the river mouth to about past decades, have become increasingly vulnerable to those
30 km upstream), the transition region (from about 30 km to events. Cities located along an estuary are at the crossroad
about 150 km upstream), and the river-dominated region (be- between the ocean and the river catchment, hence particu-
yond 150 km upstream). Thus, the local water management larly vulnerable (Herdman et al., 2018; Vitousek et al., 2017;
can define proper mitigation for handling compound flood- Zhang and Liu, 2017).
ing hazards along the riverbanks by using this zoning cate- One of the most significant drivers that can trigger com-
gory. The model also successfully reproduced a compound pound flooding over a coastal area is storm surges (Herdman
et al., 2018; Zijl et al., 2013). A storm surge is defined as the

Published by Copernicus Publications on behalf of the European Geosciences Union.

2742 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

difference between the observed water level and the expected future, a hazard assessment study in Indonesian low-lying
water level that results from tidal dynamics in a coastal area. coastal areas (such as deltas) has become critical for Indone-
A storm surge effect on usual tidal dynamics is the altered sian water management authorities. One area in Indonesia
timing of high and low water through non-linear processes that is vulnerable to coastal or even compound flooding is
(Zijl et al., 2013). A storm surge is composed of low- and the Kapuas River delta. In this delta, flooding events happen
high-frequency components (Spicer et al., 2019). The former more than once a year (Wells et al., 2016). Unfortunately,
modifies the non-tidal water level, and the latter represents there is no previous study addressing the process that under-
a tide–surge interaction during the event. A storm surge is lies the flood events in there. The previous study (Hidayat
generally quantified by the skew surge (Giloy et al., 2019), et al., 2014) successfully evaluated the inundation frequency,
which is a tidal cycle average measurement – the difference but it only did so for the upstream area and did not evaluate
between the observed height and the expected height. It re- the underlying process. Therefore, this study attempts to fill
flects the level of surge generated over a tidal cycle. To mit- in the gap and provide the first compound flood assessment
igate compound flooding hazards in a coastal area, a storm in the area.
surge is a critical component that should always be taken into This paper investigates the interaction between tides,
account. storm surges, and river discharges in the Kapuas River down-
Several factors must be considered regarding the assess- stream and its surrounding area. We use a 2D hydrody-
ment of compound flooding risks in a river delta. The first namic model to simulate the interaction between these driv-
factor is the coincidence of the sources (Herdman et al., ing forces and their effects on the compound flooding in the
2018), which means there is a possibility for the delta to re- Kapuas River delta, particularly in Pontianak. Then, we cre-
ceive an extreme flow from the upstream, while, at the same ate a detailed flood assessment, determine the area’s extent,
time, there could be an intense surge occurring in the tide or and calculate the inundation depth. As a case study, we im-
excessive rainfall over the coastal area. The second factor is plemented the model explicitly to a compound flooding event
the dependency and interdependency of the sources, indicat- on 29 December 2018.
ing whether the interaction between these sources (extreme
flows, tides, and excessive rainfall) could significantly impact
the inundation processes (Bilskie and Hagen, 2018; Herdman
et al., 2018; Santiago-Collazo et al., 2019). Other important 2 Materials and methods
factors include the vegetation properties along the riverbanks
that resist the flow of water, the vegetative properties over 2.1 Area of interest
the estuary that reduce the momentum transfer of wind, the
landscape characteristics of the coastal area, and how they The Kapuas River flows from the center of the island of Bor-
interacted with each other (Twilley et al., 2016). neo (Indonesia) toward the Karimata Strait on the western
Flooding events in a delta area can be simulated and as- coast (Fig. 1). The river is one of the longest island rivers in
sessed using hydrodynamic models (Deb and Ferreira, 2017; the world, with a length of about 1143 km (Goltenboth et al.,
Olbert et al., 2017; Patel et al., 2017). The most useful mod- 2006). The Kapuas River basin is located on the Equator with
els are those that can seamlessly simulate the processes oc- high air temperature and humidity throughout the year. The
curring along the land–sea continuum, which corresponds to river basin spreads over about 8.28 × 104 km2 , with about
the area encompassing the coast, the estuary, and the river 66.7 % of it consisting of forests (Wahyu et al., 2010). The
channels. Such models allow for a study of past flooding topography of the river comprises hills over its upstream and
events and an assessment of flood mitigation strategies (Vu plains over its downstream. The river flows into the Karimata
and Ranzi, 2017). However, realistic input variables and forc- Strait through five major branches (MacKinnon et al., 1996).
ing are needed to create an accurate physics-based hydrody- The Kapuas Kecil is the second largest distributary of the
namic model. Since the processes that drive compound flood- Kapuas. The river branch starts from a Kapuas branch in the
ing are three-dimensional, a 3D model is the most appropri- district of Rasau Jaya and ends at the estuary area in the dis-
ate tool to evaluate the event. However, the area of interest trict of Jungkat. In the middle of its streamflow (about 20 km
represents a well-mixed and relatively shallow waterbody. before the river mouth), the Landak tributary joins the Ka-
Therefore, we applied a 2D barotropic solution to reduce puas Kecil. The junction is located in the city of Pontianak,
computational costs (Huybrechts et al., 2010; Néelz, 2009). the capital of the province of West Kalimantan, Indonesia,
As an archipelago country with about 100 000 km of coast- placing the urban area at the highest risk of flooding among
lines, Indonesia is faced with significant coastal flooding other areas along the Kapuas riverbanks.
risks (World Meteorological Organization, 2019). Indonesian Pontianak is the most populated urban area on the west-
coastal flood hazards are classified as high (GFDRR, 2022). ern coast of the island of Borneo, with a population of about
It means that the potentially damaging waves are expected to 600 000 people that keeps increasing. The city is located on
flood the coasts at least once every 10 years. Based on this low-lying land and has 61 canals spread across the area. Most
risk and concerns about the impact of climate change in the of these canals flow in the Kapuas Kecil River (Pemerintah

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2743

Figure 1. The region of interest includes the Karimata Strait, the Kapuas River and its estuary, and the city of Pontianak, whose limits are
shown in green. The unstructured mesh is made of triangles whose resolution ranges from 50 m over the river and estuary to 1 km near the
coastlines and 10 km in the middle of the Karimata Strait. The mesh is composed of 206 359 triangular elements. Red dots represent the
weather station’s locations, and blue dots represent the tidal validation points (on the left is the river mouth; on the right is the Pontianak
Maritime Meteorological Station observation point, where the weather and the tidal validations points are superimposed). Blue triangles
depict the river boundary where discharges are imposed in the model. The background map was retrieved from OpenStreetMap (https:
//planet.osm.org, OpenStreetMap contributors, 2021). © OpenStreetMap contributors 2017. Distributed under the Open Data Commons
Open Database License (ODbL) v1.0.

 
Kota Pontianak, 2021). As a consequence of its geographical ∂U UU
+∇ · + f ez × U − ∇ · (υ∇U )
situation, the city often experiences inundations. ∂t H
Cd 1 H
= αgH ∇(H − h) − 2 |U |U + τwind − ∇patm , (2)
2.2 Hydrodynamic model H ρ ρ

The interplay between the river discharges, the tides, and where U = H u is the horizontal transport, H is the water col-
wind surges from the sea and their effect on the inundation umn height, h is the bathymetry, t is the time, and u = (u, v)
processes in Pontianak is investigated by using the Second- is the depth-averaged horizontal velocity and where α is a
generation Louvain-la-Neuve Ice-ocean Model (SLIM, https: constant that is set to zero over dry elements and one over wet
//www.slim-ocean.be/, last access: 3 April 2021). The model elements (Le et al., 2020), ∇ is the horizontal gradient oper-
equations are discretized with the discontinuous Galerkin fi- ator, g = 9.81 m s−2 is the gravitational acceleration, Cd is
nite element method. The model uses an unstructured mesh the bulk drag coefficient, f is the Coriolis parameter, ez is
whose resolution can vary in space. The model has success- the vertical unit vector pointing upward, υ is the horizontal
fully been applied to several areas, such as the Great Barrier eddy viscosity, ρ is the water density, τwind is the wind stress,
Reef (Lambrechts et al., 2010), the Mahakam delta (Pham and ∇patm is the atmospheric pressure gradient. Here, the
Van et al., 2016), the Scheldt estuary (Gourgue et al., 2013), wind stress (τwind ) was computed with the Smith and Banke
and the Columbia River (Vallaeys et al., 2018). (1975) formula for the wind speed of less than 20 m s−1 and
Here, we use the wetting–drying barotropic version of was computed with the Moon et al. (2007) formula for wind
SLIM that solves the conservative form of the shallow-water speed higher than 20 m s−1 .
equations (SWEs): The model equations are solved using a wetting–drying
algorithm with an implicit time-stepping scheme. With this
∂H algorithm, a mesh element can be defined as wet or dry. It
+ ∇ · U = 0, (1) requires a procedure that guarantees that the water column
∂t

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022

2744 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

Figure 2. Bathymetry map, with positive values meaning under the mean sea level.

height (H ) is always positive at the end of the time step to Here, we generate a multi-scale mesh of the entire do-
solve the equations. To achieve that, we set first a thresh- main by using the mesh-generation algorithm of Remacle
old h∗ that indicates the water height below which an ele- and Lambrechts (2018). The mesh consists of 206 359 tri-
ment is assumed to be dry (here h∗ = 0.1 m). Then, we define angular elements. Fine mesh elements are used to accurately
a parameter representing the total water column height on an represent inundation events over the land area of Pontianak
element as a combination of maximum water level and min- and along the Kapuas riverbanks, while the coarser mesh ele-
imum bathymetry: s = max(η) + min(h). When s is smaller ments are used far away from the delta. The mesh resolution
than h∗ , gravity will be canceled on the element so that the ar- reaches 50 m along the Kapuas riverbanks and over the city
tificial gradient of surface elevation would not remove water of Pontianak. It decreases to 10 km in the middle of the Kari-
from an already-dry element (α = 0). On the contrary, if s is mata Strait (Fig. 1).
greater than h∗ , then the gravitational force will be preserved The bathymetry is created from a combination of three
as usual (α = 1). The transition between α = 0 and α = 1 different data sources. The first is the river and estu-
occurs smoothly. More details about this procedure can be ary bathymetry obtained from the Indonesian Navy (Käst-
found in Le et al. (2020). ner et al., 2019) with a 100 m × 100 m grid resolution.
The second is the Karimata Strait bathymetry, obtained
2.3 Model setup from BATNAS (Badan Informasi Geospasial, 2018a) with
a 180 m × 180 m grid resolution. The last is the digital el-
Before creating the model, we first define a domain covering evation model from DEMNAS (Seamless Digital Elevation
both the ocean and the Kapuas River delta. Since we focused Model dan Batimetri Nasional; Badan Informasi Geospasial,
on evaluating the impact of tide–surge–discharge interaction 2018b) with 0.27 arcsec resolution. The Karimata Strait
on extreme water levels along the Kapuas Kecil branch (par- bathymetry shows that the strait is shallow (less than 100 m)
ticularly in Pontianak), we leave out several distributaries and relatively flat (Fig. 2). On the other hand, the bathymetry
that may not significantly influence that dynamics, such as of the river is more heterogeneous. The river is shallow in the
the southern Kubu branch. Next, we mesh the domain, set the estuary but deeper in the middle stream. The depth ranges
bathymetry and the bulk drag coefficient, define the bound- from 1 m (in the estuary area) to 40 m in the middle stream
ary conditions (surface elevation and velocity), and finally area (Kästner et al., 2017). The Kapuas Kecil River, which
impose some forcings. After the model is created and run flows through Pontianak, has a depth that decreases from
successfully, we validate the results using observational data. 15 m in the middle stream to 1 m in the river mouth.

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2745

Figure 3. Maximum daily wind speed from ECMWF ERA5 compared to observations at the Kapuas Kecil River mouth. During strong wind
events, the ERA5 data tend to underestimate the actual wind speed.

The wind velocity and the atmospheric pressure data are we imposed the discharge of the Kapuas River and the Lan-
the ERA5 reanalysis dataset obtained from the European dak River. Since there are no observational data for both
Centre for Medium-Range Weather Forecasts (ECMWF). rivers, we set constant discharges for the simulations in some
The data have a spatial resolution of 31 km (Hersbach scenarios. However, to reconstruct the observed inundation
et al., 2020), while the temporal resolution is hourly and (discussed in the case study), we imposed the discharges ob-
available at 137 vertical levels (0 to 80 km from the sur- tained from the Global Flood Monitoring System (GFMS).
face). The model wind stress parametrization uses the wind The GFMS estimated the global discharge based on Tropical
speed at 10 m height. Unfortunately, compared with ob- Rainfall Measuring Mission (TRMM) Multi-satellite Precip-
servational data from the Stasiun Klimatologi Mempawah itation Analysis and Global Precipitation Measurement data
(http://iklim.kalbar.bmkg.go.id, last access: 3 April 2021), (Wu et al., 2014). The Kapuas River discharge is bigger than
measured at 10 m above the surface, there are clear differ- the Landak River discharge (Fig. 4). At the open boundaries
ences in amplitude. The observed wind velocity is more sig- in the ocean, we prescribed the tidal elevation and current of
nificant than the wind velocity from ERA5 during a wind 15 harmonics from the global tide model dataset, the OSU
surge (Fig. 4). Therefore, in the case study, we adjust the (Oregon State University) TPXO (TOPEX/Poseidon) Tide
magnitude of the wind input data (ERA5) during the wind Models TPXO9-atlas (Egbert and Erofeeva, 2002). We also
surge event. We multiplied the wind magnitude with a ratio prescribed global ocean circulation from HYCOM (HYbrid
between both peaks (the observed and ERA5 data). Coordinate Ocean Model; Chassignet et al., 2007) at these
Next, we set the bulk drag coefficient to 2.5 × 10−3 over boundaries.
the ocean and 1.9 × 10−2 (average bulk drag during Decem- To evaluate the model performance, we performed the
ber 2018, since it is slightly increasing with the river dis- tidal analysis on the simulated surface elevation and com-
charge) over the river part (Kästner et al., 2018). Both co- pared it to the observed water level at the tide gauges. The
efficients correspond to a sandy bed. Over the dry area, the analysis was conducted using the Unified Tidal Analysis and
drag coefficient is determined by two types of land cover (ur- Prediction (UTide) Python package (Codiga, 2011). The tidal
ban and non-urban area), as defined in the Copernicus Global analysis assumes that a tidal signal is a linear combination
Land Service land cover map (Buchhorn et al., 2020). For the of multiple sinusoidal signals associated with its astronom-
urban area, the bottom drag coefficient is 2.05 (Hashimoto ical forcing. Each of these signal components has a differ-
and Park, 2008). Over the non-urban area, it is set to 2.0, ent phase, amplitude, and temporal frequency. The analysis
which corresponds to dense vegetation (Li and Busari, 2019). aims to obtain tidal harmonic constituents. Since we only
Then, we imposed discharge at the rivers and tides at the ran the model for a limited time duration (in order to min-
open ocean for the open boundaries. Upstream of the rivers, imize the computational cost), it is impossible to extract all

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022

2746 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

Figure 5. Observed and simulated tidal constituents (amplitude and

phase) at the Kapuas Kecil River mouth (0.0617◦ N, 109.1751◦ E)
during December 2018.

how well the model’s performance represents the observed

data. NSE = 1 represents a perfect model; NSE = 0 depicts
a model with a predictive skill the same as the mean of the
observed data; and NSE < 0 informs that the mean of the ob-
served data is a better predictor than the model output. The
latter coefficient indicates the average of the difference in the
peaks between both time series. Both of these coefficients
were calculated using the library hydroGOF (goodness of fit)
Figure 4. River discharge of (a) the Landak (prescribed in R (Zambrano-Bigiarini, 2020).
at 0.0282◦ S, 109.445◦ E) and (b) the Kapuas (prescribed In order to better investigate the inundation extent and
at 0.3623◦ S, 109.6394◦ E) retrieved from Wu et al. (2014).
depth over the city of Pontianak, we used the HEC-RAS
model (Hydrologic Engineering Center River Analysis Sys-
tem; U.S. Army Corps of Engineers Hydrologic Engineer-
tidal constituents from the model’s signal output. As a rule of
ing Center, 2022). We firstly created a second mesh to cover
thumb to select the appropriate tidal constituent, we used the
the computational domain over the city and used water lev-
Rayleigh criterion (Godin, 2015). Based on this criterion, we
els from the SLIM output model as the boundary conditions
obtained nine tidal constituents, i.e., K1, O1, Q1, P1, and J1
(Fig. 12). The second mesh has a 10 m resolution; hence it
(as the representation of the diurnal components); M2 and
can better represent canals that spread in the city. Like the
S2 (as the representation of the semidiurnal components);
SLIM model, the HEC-RAS model also solves the shallow-
and MK3 and MO3 (as the representation of the third-diurnal
water equations (SWEs).
constituents).
To perform tidal analysis, water level signals were ex-
tracted and analyzed at two points. The first point is located 3 Results
at the river mouth (0.0617◦ N, 109.1751◦ E), and the second
point (0.020431◦ S, 109.33852◦ E) is located along the Ka- 3.1 Model validation
puas Kecil River within the city of Pontianak (Fig. 1). The
second point is located about 20 km from the river mouth. At The validation results show a good agreement between the
the first point, the model’s output signal is compared with the model output and the ina-COAP’s tidal signal at the Kapuas
tidal signal from the Indonesian Consortium of Oceanic and Kecil River mouth (0.0617◦ N, 109.1751◦ E), with a coeffi-
Atmospheric Prediction (ina-COAP) (Pusat Riset Kelautan, cient of determination of R 2 = 0.95 and RMSE = 0.09 m. At
2021). Meanwhile, for the second point, the model’s signal this point, the model produces a tidal signal that has con-
output is compared to the observational data collected by the stituent attributes (phase and amplitude) similar to the ina-
Pontianak Maritime Meteorological Station (PMMS). COAP’s tidal signal. The diurnal components explain about
To quantify the agreement between the simulated wa- 90.69 % of the total variances of the estuary’s tidal dynam-
ter level and observation, we calculated the Nash–Sutcliffe ics, while the semidiurnal components only account for about
efficiency (NSE) and root mean square error (RMSE) be- 9.31 % (see Fig. 5 and Appendix Table A1 for more detail).
tween both time series. The former coefficient measures Next, the tidal analysis result at the observation point in Pon-

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2747

files no longer increase along with the upstream discharge’s

increases. In fact, the closer to the river mouth, the lower the
gap between the MXWL profiles. Then, all MXWL profiles
converge at about 30 km before the river mouth. We define
the part of the river as a transition region. Meanwhile, at
about 30 km from the river mouth, the discharge variations
have almost no influence on the MXWL. In this area, the
tides control the MXWL, so we define this river part as a
tidally dominated region.

3.3 The impact of wind surges on the river’s maximum

water level

To evaluate the storm surge effect, we consider wind veloc-

Figure 6. Observed and simulated tidal constituents (amplitude and ity scenarios (Fig. 8). For each scenario, we multiplied the
phase) in the middle of Pontianak (0.020431◦ S, 109.33852◦ E) dur- wind speed over the domain during the inundation event by
ing December 2018. a constant value ranging from 1 to 5 (Table A3). Then, we
evaluated their impact on the MXWL profiles along the Ka-
puas Kecil River stream. Based on the profiles (Fig. 9), we
tianak (0.020431◦ S, 109.33852◦ E) also shows a good agree- thus know that the increasing wind speed beyond 16 m s−1
ment between the model output and observation with a coef- (the highest wind speed in scenario Wind_2x) changes the
ficient of determination of R 2 = 0.83 and RMSE = 0.14 m. MXWL distribution profile. The stronger the wind speed, the
The model output signal also has similar constituents at- higher the MXWL. On the contrary, if the wind speed is less
tributed to the PMMS observational water level signal (Fig. 6 than 16 m s−1 , the MXWL profiles remain the same.
and Appendix Table A2). At this point, the diurnal compo- Furthermore, the storm surge strongly dominates the
nents explain about 89.6 % of the total variances of the es- MXWL from the river mouth to about 30 km up the river.
tuary’s tidal dynamics, which is lower from the river mouth. If the wind speed exceeds 20 m s−1 (the highest wind speed
Then, the semidiurnal components account for about 9.9 %, in scenario Wind_2.5x), the MXWL profile will increase lin-
and the shallow-water terdiurnal components explain the early due to the wind speed. However, at about 30 km from
rest. The analysis at this second point also confirms that the the river mouth, the MXWL profile drops for the Wind_4x
model successfully reproduces the observed hydrodynamics. and the Wind_5x scenarios. The decline in the MXWL pro-
file could be due to the fact that the water levels within
3.2 The impact of river discharges on the river’s
the river are higher than its riverbanks and then outflows to
maximum water level
the floodplain. The outflow to floodplains reduced the wa-
The impact of the Kapuas River upstream discharge on its ter level in the main river. Then, the MXWL profiles of all
downstream hydrodynamics is assessed by imposing differ- scenarios with wind speed of less than 24 m s−1 (the high-
ent discharge values, ranging from 3000 to 9000 m3 s−1 (the est wind speed in the Wind_3x scenario) converged at about
highest discharge based on the simulated data in 2018; see 150 km upstream. This convergence matches with the previ-
Fig. 4 and Table A3). Only 17 % of these values will divert ous study of Kästner et al. (2019), which found that at this
to the Kapuas Kecil branch, while 83 % of it continues to point, the admittance of the tidal propagation upstream has
flow through the Kapuas Besar (Kästner and Hoitink, 2019). a knickpoint, where dumping strongly increases. This point
Meanwhile, the Landak River upstream discharge is set to also confirmed that it is the upstream boundary of the transi-
300 m3 s−1 in all scenarios. The tides used in the ocean part tion zone. However, if the wind speed is greater than 24 m s−1
were retrieved from TPXO for December 2018, with a tidal (the Wind_4x and the Wind_5x scenario), the MXWL pro-
range of 1.8 m at the river mouth. The maximum water level files remain slightly higher (do not converge). It means the
(MXWL) at different points along the Kapuas River stream is surge impact is still traveling further upstream.
then calculated based on these different discharge scenarios Besides the wind speed, the flood duration and flood ex-
(Fig. 7). The MXWL profiles show that the river discharge tent along the river are also influenced by the storm duration
dominates the maximum water level from the river upstream (Höffken et al., 2020). Figure 10 shows that the impact of the
to about 150 km before the river mouth. The MXWL pro- wind surge duration on the maximum water level is not sig-
file increases along with the discharge’s increases, so the nificant. Still, it makes the flooding event longer – backwater
water level fluctuation in this stream part can be said to be that comes from the river mouth upstream stays longer inland
river-dominated. Within about 150 km from the river mouth, before flowing back to the ocean.
the MXWL profiles between the “with-tides” and “without-
tides” scenarios show a different pattern. The MXWL pro-

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022

2748 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

Figure 7. River discharge affecting the maximum water level, with measurements for the (a) tidally dominated region, (b) transition region,
and (c) river-dominated region. The dashed vertical lines indicate the limits between the different regions. The arrow shows the river stream
that flows inside Pontianak, where the population is the densest in the domain.

Figure 8. Wind direction and magnitude over the domain on 29 December 2018 at 04:00 UTC.

3.4 A case study: analysis of the flood event on the Kapuas and Landak rivers had discharges of 4400 and
29 December 2018 502 m3 s−1 , respectively. At the river mouth, the tidal range
reached 1.8 m.
Based on the water level observed by the Pontianak Mar- In order to investigate the main drivers of the event, we
itime Meteorological Station (see Fig. 1 for the location), simulated the hydrodynamical process along the land–sea
on 29 December 2018, there was a significant increase in continuum for the full month of December 2018. We sim-
water levels from 05:20 to 07:50 UTC. The maximum wa- ulated the hydrodynamical process without and with storm
ter level reached 2.8 m and produced a significant inunda- surge scenarios when the expected tidal range during the
tion in the city of Pontianak. The 2.8 m water level refer- event is 1.8 m. Since there was no observed discharge up-
ence is the lowest astronomical tide (LAT). It corresponds stream during the date, we imposed the river discharge re-
roughly to 1.8 m a.m.s.l. (above mean sea level) and 0.7 m
above the highest astronomical tide (HAT). During the event,

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2749

Figure 9. Wind surge affecting the maximum water level, with measurements for the (a) tidally dominated region, (b) transition region, and
(c) river-dominated region. The arrow shows the river stream that flows inside Pontianak.

Figure 10. Wind surge duration affecting the increasing water level in Pontianak. Please note that the date format used in this figure and
below is year-month-day.

trieved from the GFMS for the Kapuas and Landak rivers again for a short moment. The storm alone is responsible for
(Table A3). a 30 cm increase in the water level during this short moment
The validation result shows that the NSE is 0.82, indicat- (light-grey box in Fig. 11). After this additional forcing ef-
ing that the model has a good performance. The RMSE of fect disappeared, the water level then decreased steeply and
its peaks is 0.04 m. These results suggest that the model cor- started following the tidal signal again.
rectly reproduces the actual hydrodynamic processes. The in- A more qualitative validation of the model has also been
undation event on 29 December 2018 is well depicted by the performed by comparing the extent of the simulated inun-
model (light-grey box in Fig. 11). The observational data pro- dation area with observations reported by the local media
file (black) shows that the water level dynamic is at the peak (Madrosid, 2018). The report mentioned some areas that
moment during the event, where its peak seems lower than were confirmed as inundated when the flooding occurred
the previous peak period. The water level dynamic is about (represented by the red dots in Fig. 13). As a result, our
to go down when suddenly a strong force pushes it to go up

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022

2750 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

Figure 11. Validation of water level at Pontianak, where the light-grey box depicts the peak of the inundation. BMKG: Badan Meteorologi,
Klimatologi, dan Geofisika (Meteorology, Climatology, and Geophysical Agency of Indonesia).

Figure 12. HEC-RAS 2D mesh with a resolution of 10 m for flood extent analysis in the city of Pontianak. WL: water level.

model output validated that these areas were flooded during flooded densely populated areas near riverbanks and propa-
the flooding event. gated much further into the city through the drainage chan-
We further investigated the inundation extent and depth nels. In the western part of the city, floods impacted ma-
over the city of Pontianak using the output of the HEC-RAS jor areas near riverbanks, while in the eastern part, the ma-
model. From the model outputs, we thus know that the ex- jor flooded areas fell in lowland areas, located between the
tent of this inundation reached lowland within the city lo- Kapuas Kecil and the Landak rivers (district of Pontianak
cated far from the riverbanks (Fig. 13). The overtopped water Timur), and portioned the lowland area in the northeast.

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2751

Figure 13. Inundation extent map and its depth during the flood event on 29 December 2018 at 06:00 UTC. The red dots represent places
that were reported as inundated during the event. The city’s building map was retrieved from OpenStreetMap (https://planet.osm.org, Open-
StreetMap contributors, 2021). © OpenStreetMap contributors 2017. Distributed under the Open Data Commons Open Database License
(ODbL) v1.0.

4 Discussion

Based on the discharge scenarios, we found that the city of

Pontianak is located in a tidally dominated region. There-
fore, the MXWL of the river in the city is not controlled by
discharges upstream anymore. Next, we also found that an
extreme eastward wind speed with a magnitude greater than
16 m s−1 (the highest wind speed in the Wind_2x scenario)
can drive the seawater into the river channels based on the
wind surge scenarios. As the storm surge continues to push
the water upstream, the pileup water in the estuary will en-
Figure 14. The total inundated area within the city of Pontianak ter the river stream and propagate over several kilometers.
during the flood event on 29 December 2018. If the water propagation is in co-occurrence with high river
discharge, the water level can hence overtop riverbanks and
lead to a compound inundation over the floodplain. However,
Next, the peak of the flooding event occurred during a short
if the wind speed is less than 16 m s−1 , it does not impact the
period. It started at 01:00 UTC, peaked at 06:00 UTC, and
MXWL profiles. Therefore, the wind speed of 16 m s−1 be-
finished at 15:00 UTC. At the peak moment, the area inun-
comes the minimum wind speed that can drive surges inside
dated was about 12 km2 (Fig. 14), in varying depths.
the river stream.
On 29 December 2018, rainfall over the city of Pontianak
was less than 7 mm so that the effect of excessive rainfall
could be ruled out in this event. Meanwhile, an intense wind

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022

2752 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

speed was observed over the coastal area for a few hours. The 5 Conclusion
radar data show cumulonimbus convective clouds formed
and moving eastward from the Karimata Strait towards the In this study, a model that depicts the Kapuas River’s hy-
land (see Appendix Fig. B1). Cumulonimbus is a cloud type drodynamic processes and its interaction with the Karimata
that could produce wind surges, tornadoes, and excessive Strait tides and wind surges has been successfully set up.
rainfall (Cotton et al., 1992). These clouds reached and cov- We simulated the hydrodynamic processes during extreme
ered the Kapuas River’s mouth on that date, from 04:30 to events using the model and assessing the impact on the maxi-
05:25 UTC. Therefore, these clouds most likely triggered a mum water level along the Kapuas River. We found that wind
storm over the coastal area with wind speed ranging from 13 surges with speeds between 16 and 24 m s−1 over the Ka-
to 21 m s−1 . puas estuary can raise the water level to 150 km upstream. If
The wind direction was oriented from the west to the east the wind speed is more than 24 m s−1 , then the surge impact
during the event. Hence, the total wind effect became the can travel further beyond this point. The stronger the wind
combination of the direct wind stress effect and the indirect speed, the further the surge impact will travel upstream. If
wave run-up effect. Consequently, it drove the water column the surge propagation upstream occurs during intermediate
from the ocean to the western coast of the island of Borneo, or high river discharges, it can significantly increase the river
where the Kapuas estuary is located. Then, the wind piled up water level and trigger an inundation over the floodplain, in-
the tidal waters inside the Kapuas River mouth. The piled- cluding the city of Pontianak.
up water then propagated upstream from the river mouth Next, based on the maximum water level profiles, we de-
and was coincidentally met with an intermediate river dis- lineate the stream along the Kapuas River into three regions.
charge. This phenomenon created an increasing water level From the river mouth to about 30 km up the river is the tidally
and caused a short-duration overflow over the floodplain. dominated region, where the river discharge levels do not im-
Therefore, using this scenario, we suggest that the main trig- pact the maximum water level anymore. From about 30 to
ger for the compound flooding, which occurred on 29 De- 150 km is the transition region, where the maximum water
cember 2018, was an interaction between a storm surge at level is influenced by the interaction between the tides, the
the estuary area and a medium discharge from the Kapuas surges, and the river discharges. Then, from about 150 km
River. upstream is the river-dominated region, where surges (with
As with all modeling studies, there are some limitations wind speed less than 24 m s−1 ) no longer impact the maxi-
related to the model that should be mentioned. Firstly, many mum water level; with- and without-tides scenarios have sim-
different compound flooding schemes possibly occur in this ilar MXWL profiles. The local water management agency
area that have not yet been simulated. Therefore, the delin- can use this zoning category in assessing and mitigating the
eation of the compound flooding risk zones based on the compound flooding hazard along the riverbanks.
MXWL proposed for the Kapuas River needs further inves- Lastly, as a case study, the factors that drive the inunda-
tigation in the future. Secondly, we have not yet imposed tion event over Pontianak on 29 December 2018 have been
excessive rainfall scenarios in the model so that the model successfully investigated. The wind surge, which occurred
only depicts inundation processes as the effect of the river in the estuary area, was concluded to be the main trigger of
discharges, tides, and wind surge. In the actual case, a sin- the flood event. This wind surge pushed seawater upstream
gle instance of excessive rainfall is enough to trigger urban and met with a medium river discharge, therefore triggering
flooding over the floodplain area. Next, the computational a short inundation event over the floodplains, especially the
domain does not cover all of the dry land over the delta. It city of Pontianak.
is only limited to Pontianak and the Kapuas Kecil riverbanks
from the city to the river mouth. Consequently, the simula-
tion might not wholly describe the inundation processes and
the real extent of the inundated area. These limitations will
be improved in future work.
Regarding the event’s impact on 29 December 2018, the
flooding occurred on a day of low precipitation. Therefore,
the residents were not aware and not prepared for such a sig-
nificant flood event. Even though it happened only for a short
period, the economic loss was severe (Madrosid, 2018). Such
loss can hopefully be avoided after the local water manage-
ment agency installs a flood warning system built based on
our proposed models.

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2753

Appendix A: Detail on the tidal constituents

The amplitude, phase, and energy of each tidal constituent

at the river mouth and in Pontianak are summarized in Ta-
bles A1 and A2.

Table A1. Tide constituents at the Kapuas Kecil River mouth (0.0617◦ N, 109.1751◦ E) in December 2018.

Tidal ina-COAP SLIM Error Rayleigh

constituent Amplitude Phase Percent energy Amplitude Phase Percent energy 1Amplitude 1Phase criterion
(m) (◦ ) (%) (m) (◦ ) (%) (m) (◦ )
K1 0.36 304.59 54.39 0.29 307.29 48.03 0.06 2.71 0.16
O1 0.26 349.21 29.56 0.22 353.95 27.30 0.04 4.75 1.09
M2 0.14 252.24 8.62 0.16 252.74 13.46 0.02 0.50 26.80
P1 0.11 278.45 4.95 0.12 286.67 7.88 0.01 8.22 2.03
Q1 0.05 202.88 1.03 0.05 185.47 1.31 0.00 17.40 24.77
J1 0.04 99.54 0.69 0.01 149.53 0.11 0.03 50.00 1.09
S2 0.04 136.68 0.76 0.06 143.77 1.75 0.01 7.10 2.03

Table A2. Tide constituents in the middle of Pontianak (0.020431◦ S, 109.33852◦ E) in December 2018.

Tidal Observation SLIM Error Rayleigh

constituent Amplitude Phase Percent energy Amplitude Phase Percent energy 1Amplitude 1Phase criterion
(m) (◦ ) (%) (m) (◦ ) (%) (m) (◦ )
K1 0.23 65.24 42.42 0.24 54.85 42.11 0.01 4.65 0.16
O1 0.18 108.39 25.33 0.20 90.81 28.71 0.02 3.64 1.09
M2 0.11 124.73 13.34 0.14 81.92 14.53 0.04 13.83 26.80
P1 0.1 78.8 10.54 0.12 28.50 9.73 0.02 35.35 2.03
Q1 0.06 299.63 4.11 0.05 279.19 1.71 0.01 7.04 24.77
J1 0.04 238.67 1.78 0.02 269.16 0.37 0.02 46.08 1.09
S2 0.04 25.4 1.54 0.06 3.22 2.21 0.02 352.17 2.03
MK3 0.02 163.32 0.52 0.03 47.35 0.46 0.01 71.94 2.20
MO3 0.01 216.89 0.42 0.01 122.99 0.17 0.00 50.98 25.85

Table A3. Scenarios used to force the model. The simulation duration is 1 whole month, with wind direction varying from 0 to 360◦ , pressure
varying from 100.5 to 101.5 kPa, and a tidal range of 1.8 m.

Scenario Wind speed Discharge Kapuas Discharge Landak

(m s−1 ) (m3 s−1 ) (m3 s−1 )
Discharge1 2–8 3 × 103 300
Discharge2 2–8 5 × 103 300
Discharge3 2–8 7 × 103 300
Discharge4 2–8 9 × 103 300
Wind_1x 2–8 3 × 103 300
Wind_1.5x 3–12 3 × 103 300
Wind_2x 4–16 3 × 103 300
Wind_2.5x 5–20 3 × 103 300
Wind_3x 6–24 3 × 103 300
Wind_4x 8–32 3 × 103 300
Wind_5x 10–40 3 × 103 300
Case study 2–21 3.3 × 103 –5 × 103 250–700

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022

2754 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

Appendix B: Local weather conditions on

29 December 2018

Figure B1. Radar data, observed by the Supadio Meteorological Station (https://kalbar.bmkg.go.id/, last access: 3 April 2021), depict the
clouds’ growth and movement (in the red box) on 29 December 2018. The clouds started to grow over the ocean and are pushed by the
wind to move eastward at 02:59 UTC. The clouds reach their maximum intensity at 05:00 UTC while located above the Kapuas estuary.
Furthermore, the clouds spread over a wider area with a weakening intensity at 07:00 UTC.

Code availability. The SLIM code and step-by-step instructions on Competing interests. The contact author has declared that neither
how to run the simulation can be accessed at https://git.immc.ucl. they nor their co-authors have any competing interests.
ac.be/slim/slim/-/wikis/home (SLIM contributors, 2022). However,
since this project has been conducted on several separate computers
(including high-performance computers – HPCs – and desktops), Disclaimer. Publisher’s note: Copernicus Publications remains
we decided to not publicly release the code of this specific project. neutral with regard to jurisdictional claims in published maps and
The code is only available to the internal SLIM team. institutional affiliations.

Data availability. The gauged water level in Pontianak for 2018, Special issue statement. This article is part of the special issue “To-
which we used in the case study, is available at https://doi.org/10. wards an understanding and assessment of human impact on coastal
5281/zenodo.5809647 (Sampurno, 2021). The data were collected marine environments”. It is not associated with a conference.
at the Pontianak Maritime Meteorological Station (PMMS), Indone-
sia. To use the data, please cite this article and the official web
page of PMMS at https://maritim.kalbar.bmkg.go.id/ (last access: Acknowledgements. The PhD fellowship of Joko Sampurno is pro-
30 May 2022). vided by the Indonesian Endowment Fund for Education (LPDP;
grant no. 201712220212183). Computational resources have been
provided by the supercomputing facilities (Calcul Intensif et Stock-
Author contributions. JS, VV, and EH conceptualized the research. age de Masse; CISM) of the Université catholique de Louvain
JS and RA curated the data. JS, VV, and EH analyzed the data. JS (UCL) and the Consortium des Équipements de Calcul Intensif
wrote the original draft of the manuscript. JS and EH reviewed and en Fédération Wallonie Bruxelles (CÉCI) funded by the Fonds de
edited the manuscript. la Recherche Scientifique de Belgique (F.R.S.-FNRS) (convention
no. 2.5020.11) and by the Walloon Region.

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2755

Financial support. This research has been supported by the Lem- Godin, G.: The resolution of tidal constituents, Int. Hydrogr.
baga Pengelola Dana Pendidikan (grant no. 201712220212183). Rev., 47, https://journals.lib.unb.ca/index.php/ihr/article/view/
23916 (last access: 30 May 2022), 2015.
Goltenboth, F., Timotius, K. H., Milan, P. P., and Margraf, J.: Ecol-
Review statement. This paper was edited by Marilaure Grégoire ogy of insular Southeast Asia: the Indonesian archipelago, edited
and reviewed by Karl Kästner and Vera Fofonova. by: Goltenboth, F., Timotius, K., Milan, P., and Margraf, J., Else-
vier B. V., https://doi.org/10.1016/B978-0-444-52739-4.X5000-
1, 2006.
Gourgue, O., Baeyens, W., Chen, M. S., de Brauwere, A.,
References de Brye, B., Deleersnijder, E., Elskens, M., and Legat,
V.: A depth-averaged two-dimensional sediment transport
Badan Informasi Geospasial: Batimetri Nasional, model for environmental studies in the Scheldt Estuary
https://tanahair.indonesia.go.id/demnas/#/batnas (last access: and tidal river network, J. Marine Syst., 128, 27–39,
14 July 2021), 2018a. https://doi.org/10.1016/j.jmarsys.2013.03.014, 2013.
Badan Informasi Geospasial: Seamless Digital Elevation Model Hashimoto, H. and Park, K.: Two-dimensional urban flood simula-
Nasional (DEMNAS), https://tanahair.indonesia.go.id/demnas/#/ tion: Fukuoka flood disaster in 1999, WIT Trans. Ecol. Envir.,
(last access: 1 January 2022), 2018b. 118, 59–67, https://doi.org/10.2495/FRIAR080061, 2008.
Bevacqua, E., Vousdoukas, M. I., Zappa, G., Hodges, K., Shepherd, Herdman, L., Erikson, L., and Barnard, P.: Storm surge prop-
T. G., Maraun, D., Mentaschi, L., and Feyen, L.: More mete- agation and flooding in small tidal rivers during events of
orological events that drive compound coastal flooding are pro- mixed coastal and fluvial influence, J. Mar. Sci. Eng., 6, 158,
jected under climate change, Commun. Earth Environ., 11, 1–11, https://doi.org/10.3390/JMSE6040158, 2018.
https://doi.org/10.1038/s43247-020-00044-z, 2020. Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A.,
Bilskie, M. V. and Hagen, S. C.: Defining Flood Zone Transitions in Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schep-
Low-Gradient Coastal Regions, Geophys. Res. Lett., 45, 2761– ers, D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Bal-
2770, https://doi.org/10.1002/2018GL077524, 2018. samo, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M.,
Buchhorn, M., Lesiv, M., Tsendbazar, N.-E., Herold, M., Ber- Chiara, G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R.,
tels, L., and Smets, B.: Copernicus Global Land Cover Flemming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger,
Layers – Collection 2, Remote Sens.-Basel, 12, 1044, L., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley,
https://doi.org/10.3390/RS12061044, 2020. S., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., Rosnay, P.,
Chassignet, E. P., Hurlburt, H. E., Smedstad, O. M., Halli- Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J.: The ERA5
well, G. R., Hogan, P. J., Wallcraft, A. J., Baraille, R., global reanalysis, Q. J. Roy. Meteor. Soc., 146, 1999–2049,
and Bleck, R.: The HYCOM (HYbrid Coordinate Ocean https://doi.org/10.1002/qj.3803, 2020.
Model) data assimilative system, J. Marine Syst., 65, 60–83, Hidayat, H., Hoekman, D., Vissers, M., Hossain, M. M., Teul-
https://doi.org/10.1016/J.JMARSYS.2005.09.016, 2007. ing, A., and Haryani, G.: Inundation Frequency Mapping of The
World Meteorological Organization: Coastal Flooding Fore- Upper Kapuas Wetlands Using Radar Imagery, in: International
cast Strengthened in Indonesia, https://public.wmo.int/en/media/ Conference on Ecohydrology (ICE), Yogyakarta, Indonesia, 10–
news/coastal-flooding-forecast-strengthened-indonesia (last ac- 12 November 2014, 250–257, 2014.
cess: 3 April 2021), 2019. Höffken, J., Vafeidis, A. T., MacPherson, L. R., and Dan-
Codiga, D. L.: Unified Tidal Analysis and Prediction, Graduate gendorf, S.: Effects of the Temporal Variability of Storm
School of Oceanography, University of Rhode Island, Narra- Surges on Coastal Flooding, Front. Mar. Sci., 7, 98,
gansett, RI, 59 pp., 2011. https://doi.org/10.3389/FMARS.2020.00098, 2020.
Cotton, W. R., Bryan, G., and van den Heever, S. C.: Cumulonim- Huybrechts, N., Villaret, C., and Hervouet, J.-M.: Comparison be-
bus Clouds and Severe Convective Storms, in: International Geo- tween 2D and 3D modelling of sediment transport: application
physics Book series, Vol. 44: Storm and Cloud Dynamics, edited to the dune evolution, in: River Flow 2010, Braunschweig, Ger-
by: Cotton, W. R. and Anthes, R. A., Academic Press, 455–592, many, 8–10 September 2010, 887–894, 2010.
https://doi.org/10.1016/S0074-6142(08)60548-3, 1992. Kästner, K., Hoitink, A. J. F., Vermeulen, B., Geertsema, T. J.,
Deb, M. and Ferreira, C. M.: Potential impacts of the and Ningsih, N. S.: Distributary channels in the fluvial to
Sunderban mangrove degradation on future coastal flood- tidal transition zone, J. Geophys. Res.-Earth, 122, 696–710,
ing in Bangladesh, J. Hydro-Environ. Res., 17, 30–46, https://doi.org/10.1002/2016JF004075, 2017.
https://doi.org/10.1016/j.jher.2016.11.005, 2017. Kästner, K., Hoitink, A. J. F., Torfs, P. J. J. F., Vermeulen, B.,
Egbert, G. D. and Erofeeva, S. Y.: Efficient inverse modeling of Ningsih, N. S., and Pramulya, M.: Prerequisites for Accu-
barotropic ocean tides, J. Atmos. Ocean. Tech., 19, 183–204, rate Monitoring of River Discharge Based on Fixed-Location
2002. Velocity Measurements, Water Resour. Res., 54, 1058–1076,
GFDRR (Global Facility for Disaster Reduction and Recovery): https://doi.org/10.1002/2017WR020990, 2018.
Think Hazard – Indonesia, https://thinkhazard.org/en/report/ Kästner, K., Hoitink, A. J. F., Torfs, P. J. J. F., Deleersni-
116-indonesia, last access: 8 January 2022. jder, E., and Ningsih, N. S.: Propagation of tides along
Giloy, N., Hamdi, Y., Bardet, L., Garnier, E., and Du- a river with a sloping bed, J. Fluid Mech., 872, 39–73,
luc, C. M.: Quantifying historic skew surges: an example https://doi.org/10.1017/JFM.2019.331, 2019.
for the Dunkirk Area, France, Nat. Hazards, 98, 869–893,
https://doi.org/10.1007/s11069-018-3527-1, 2019.

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022

2756 J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges

Kästner, K. and Hoitink, A. J. F.: Flow and Suspended Sediment Di- Pusat Riset Kelautan, Kementerian Kelautan dan Perikanan RI:
vision at Two Highly Asymmetric Bifurcations in a River Delta: Prediksi Pasang Surut, https://pusriskel.litbang.kkp.go.id/index.
Implications for Channel Stability, J. Geophys. Res.-Earth, 124, php/en/data/prediksi-pasang-surut, last access: 5 April 2021.
2358–2380, https://doi.org/10.1029/2018JF004994, 2019. Remacle, J. F. and Lambrechts, J.: Fast and robust mesh generation
Lambrechts, J., Humphrey, C., McKinna, L., Gourge, O., Fabricius, on the sphere – Application to coastal domains, Comput. Aided
K. E., Mehta, A. J., Lewis, S., and Wolanski, E.: Importance of Design, 103, 14–23, https://doi.org/10.1016/j.cad.2018.03.002,
wave-induced bed liquefaction in the fine sediment budget of 2018.
Cleveland Bay, Great Barrier Reef, Estuar. Coast. Shelf S., 89, Sampurno, J.: Observed Kapuas River water level in 2018, Zenodo
154–162, https://doi.org/10.1016/j.ecss.2010.06.009, 2010. [data set], https://doi.org/10.5281/zenodo.5809647, 2021.
Le, H.-A., Lambrechts, J., Ortleb, S., Gratiot, N., Deleersnijder, Santiago-Collazo, F. L., Bilskie, M. V., and Hagen, S. C.: A
E., and Soares-Frazão, S.: An implicit wetting-drying algorithm comprehensive review of compound inundation models in low-
for the discontinuous Galerkin method: application to the Tonle gradient coastal watersheds, Environ. Model. Softw., 119, 166–
Sap, Mekong River Basin, Environ. Fluid Mech., 20, 923–951, 181, https://doi.org/10.1016/J.ENVSOFT.2019.06.002, 2019.
https://doi.org/10.1007/s10652-019-09732-7, 2020. SLIM contributors: SLIM code, GitHub [code], https://git.immc.
Li, C. and Busari, A. O.: Hybrid modeling of flows ucl.ac.be/slim/slim/-/wikis/home, last access: 30 May 2022.
over submerged prismatic vegetation with different Smith, S. D. and Banke, E. G.: Variation of the sea surface drag
areal densities, Eng. Appl. Comp. Fluid, 13, 493–505, coefficient with wind speed, Q. J. Roc. Meteor. Soc., 101, 665–
https://doi.org/10.1080/19942060.2019.1610501, 2019. 673, https://doi.org/10.1002/QJ.49710142920, 1975.
MacKinnon, K., Hatta, G., Mangalik, A., and Halim, H.: The ecol- Spicer, P., Huguenard, K., Ross, L., and Rickard, L. N.: High-
ogy of Kalimantan, The Ecology of Indonesia Series, Vol. III, Frequency Tide-Surge-River Interaction in Estuaries: Causes and
Oxford University Press, ISBN 0-945971-73-7, 1996. Implications for Coastal Flooding, J. Geophys. Res.-Oceans,
Madrosid: Cerita Warga, Detik-detik Banjir Rob 124, 9517–9530, https://doi.org/10.1029/2019JC015466, 2019.
Melanda Kota Pontianak, Trib. Pontianak, Twilley, R. R., Bentley, S. J., Chen, Q., Edmonds, D. A., Ha-
https://pontianak.tribunnews.com/2018/12/29/ gen, S. C., Lam, N. S. N., Willson, C. S., Xu, K., Braud,
cerita-warga-detik-detik-banjir-rob-melanda-kota-pontianak D. W., Hampton Peele, R., and McCall, A.: Co-evolution of
(last access: 5 April 2021), 2018. wetland landscapes, flooding, and human settlement in the
Moftakhari, H. R., AghaKouchak, A., Sanders, B. F., Feldman, D. Mississippi River Delta Plain, Sustain. Sci., 11, 711–731,
L., Sweet, W., Matthew, R. A., and Luke, A.: Increased nui- https://doi.org/10.1007/s11625-016-0374-4, 2016.
sance flooding along the coasts of the United States due to sea U.S. Army Corps of Engineers Hydrologic Engineering Center:
level rise: Past and future, Geophys. Res. Lett., 42, 9846–9852, The Hydrologic Engineering Center’s River Analysis System
https://doi.org/10.1002/2015GL066072, 2015. (HEC-RAS), https://www.hec.usace.army.mil/software/hec-ras/,
Moon, I. J., Ginis, I., Hara, T., and Thomas, B.: A Physics-Based Pa- last access: 8 January 2022.
rameterization of Air–Sea Momentum Flux at High Wind Speeds Vallaeys, V., Kärnä, T., Delandmeter, P., Lambrechts, J., Bap-
and Its Impact on Hurricane Intensity Predictions, Mon. Weather tista, A. M., Deleersnijder, E., and Hanert, E.: Discon-
Rev., 135, 2869–2878, https://doi.org/10.1175/MWR3432.1, tinuous Galerkin modeling of the Columbia River’s cou-
2007. pled estuary-plume dynamics, Ocean Model., 124, 111–124,
Néelz, S.: Desktop review of 2D hydraulic modelling packages, En- https://doi.org/10.1016/j.ocemod.2018.02.004, 2018.
vironment Agency, Bristol, 2009. Vitousek, S., Barnard, P. L., Fletcher, C. H., Frazer, N., Erik-
Olbert, A. I., Comer, J., Nash, S., and Hartnett, M.: High-resolution son, L., and Storlazzi, C. D.: Doubling of coastal flooding fre-
multi-scale modelling of coastal flooding due to tides, storm quency within decades due to sea-level rise, Sci. Rep.-UK, 7,
surges and rivers inflows. A Cork City example, Coast. Eng., 1–9, https://doi.org/10.1038/s41598-017-01362-7, 2017.
121, 278–296, https://doi.org/10.1016/j.coastaleng.2016.12.006, Vu, T. T. and Ranzi, R.: Flood risk assessment and coping capacity
2017. of floods in central Vietnam, J. Hydro-Environ. Res., 14, 44–60,
Patel, D. P., Ramirez, J. A., Srivastava, P. K., Bray, M., and https://doi.org/10.1016/j.jher.2016.06.001, 2017.
Han, D.: Assessment of flood inundation mapping of Surat Wahyu, A., Kuntoro, A., and Yamashita, T.: Annual and Seasonal
city by coupled 1D/2D hydrodynamic modeling: a case ap- Discharge Responses to Forest/Land Cover Changes and Climate
plication of the new HEC-RAS 5, Nat. Hazards, 89, 93–130, Variations in Kapuas River Basin, Indonesia, J. Int. Dev. Coop.,
https://doi.org/10.1007/s11069-017-2956-6, 2017. 16, 81–100, https://doi.org/10.15027/29807, 2010.
Pemerintah Kota Pontianak: Kondisi Geografis Kota Pontianak, Wells, J. A., Wilson, K. A., Abram, N. K., Nunn, M., Gaveau, D.
https://www.pontianakkota.go.id/tentang/geografis, last access: L. A., Runting, R. K., Tarniati, N., Mengersen, K. L., and Mei-
5 April 2021. jaard, E.: Rising floodwaters: Mapping impacts and perceptions
OpenStreetMap contributors: Planet dump, https://planet.osm.org, of flooding in Indonesian Borneo, Environ. Res., 11, 064016,
last access: 3 April 2021. https://doi.org/10.1088/1748-9326/11/6/064016, 2016.
Pham Van, C., de Brye, B., Deleersnijder, E., F Hoitink, A. Wu, H., Adler, R. F., Tian, Y., Huffman, G. J., Li, H., and Wang, J.:
J., Sassi, M., Spinewine, B., Hidayat, H., and Soares-Frazão, Real-time global flood estimation using satellite-based precipita-
S.: Simulations of the flow in the Mahakam river-lake- tion and a coupled land surface and routing model, Water Resour.
delta system, Indonesia, Environ. Fluid Mech., 16, 603–633, Res., 50, 2693–2717, https://doi.org/10.1002/2013WR014710,
https://doi.org/10.1007/s10652-016-9445-4, 2016. 2014.

Biogeosciences, 19, 2741–2757, 2022 https://doi.org/10.5194/bg-19-2741-2022

J. Sampurno et al.: Modeling interactions between tides, storm surges, and river discharges 2757

Zambrano-Bigiarini, M.: hydroGOF: Goodness-of-fit functions for Zijl, F., Verlaan, M., and Gerritsen, H.: Improved water-level fore-
comparison of simulated and observed hydrological time series casting for the Northwest European Shelf and North Sea through
(R package version 0.4-0), The R Foundation for Statistical Com- direct modelling of tide, surge and non-linear interaction,
puting, Vienna, Austria, https://cran.r-project.org/web/packages/ Ocean Dynam., 63, 823–847, https://doi.org/10.1007/s10236-
hydroGOF/index.html (last access: 30 May 2022), 2020. 013-0624-2, 2013.
Zhang, J. and Liu, H.: Modeling of waves overtopping and flooding
in the coastal reach by a non-hydrostatic model, in: Procedia IU-
TAM, 126–130, https://doi.org/10.1016/j.piutam.2017.09.019,
2017.

https://doi.org/10.5194/bg-19-2741-2022 Biogeosciences, 19, 2741–2757, 2022